1·www.linkedin.com
The Log: Real-Time Data's Unifying Abstraction (LinkedIn)
Why I added thisWhy commit and change logs are the backbone of modern real-time data pipelines and distributed systems.
A CURATED READING LIST
Database Engineering: Replication, Storage & Specialized Systems
A reading path through 12 articles, papers and reference resources on database internals and distributed data systems. Explore logs and change data capture, MySQL replication, caching, Cassandra and Dynamo architecture, SSTable compaction, time-series systems (Gorilla, Goku and AresDB), and graph systems (TAO and Neo4j). Inspired by Tannika Majumder’s database reading collection, linked below as the original source. Reading order and curator notes help connect the resources into a learning path. Start with The Log, then MySQL replication and Netflix DBLog. Continue through caching, distributed storage and compaction before comparing specialized time-series and graph systems. Intended for backend engineers familiar with SQL and basic distributed-systems concepts. Several readings describe historical systems; use them to study design trade-offs. The Cassandra 3.0 documentation is archived and should be read as historical reference, rather than current setup guidance.
Based on resources shared in the original source.
Curated by SiddharthUpdated Nov 2, 2026
Start hereThe Log: Real-Time Data's Unifying Abstraction (LinkedIn) ↗
Read at your own pace. Save this list to keep track of your progress.
2·dev.mysql.com
MySQL 8.0 Reference Manual / Replication
Why I added thisExplains how MySQL synchronizes and replicates data across multiple servers for high availability and scalability.
3·netflixtechblog.com
DBLog: A Generic Change-Data-Capture Framework
Why I added thisNetflix's real-time framework for tracking database changes and building low-latency data pipelines.
4·engineering.fb.com
Scaling memcache at Facebook
Why I added thisHow Facebook scaled memcached to billions of reads and writes while balancing speed, consistency, and cache invalidation.
5·docs.datastax.com
Apache Cassandra | Architecture in brief
Why I added thisArchived Cassandra 3.0 documentation explaining the peer-to-peer model, replication and partitioning. Read it for architectural foundations and historical context; consult current Cassandra documentation for configuration and operational guidance.
6·www.allthingsdistributed.com
Amazon's Dynamo
Why I added thisThe classic paper on Dynamo's highly available, eventually consistent key-value storage at Amazon.
7·github.com
Scylla DB | SSTable compaction and compaction strategies
Why I added thisHow ScyllaDB and Cassandra manage on-disk data with compaction strategies to optimize reads and writes.
8·medium.com
Goku: Building a scalable and high performant time series database system
Why I added thisHow Pinterest engineered a scalable, high-performance time series database for analytics workloads.
9·www.uber.com
Introducing AresDB: Uber’s GPU-Powered Open Source, Real-time Analytics Engine
Why I added thisUber's columnar, GPU-accelerated real-time analytics engine for massive time-series and event data.
10·www.vldb.org
Gorilla: A Fast, Scalable, In-Memory Time Series Database
Why I added thisA fast, in-memory, highly compressed time-series database built by Facebook for storing monitoring data.
11·engineering.fb.com
TAO: The power of the graph
Why I added thisFacebook's TAO system for serving the social graph at massive scale.
12·neo4j.com
Native vs. non-native graph database
Why I added thisThe difference between native graph databases like Neo4j and non-native systems, and why architecture matters for graph workloads.
ReadingTrackA little more read. A little more remembered.